Determining likelihood of an adverse health event based on various physiological diagnostic states

Abstract

Techniques for determining a likeliness that a patient may incur an adverse health event are described. An example technique may include utilizing a probability model that uses as evidence nodes various diagnostic states of physiological parameters, which may include one or more subcutaneous impedance parameters. The probability model may include a Bayesian Network that determines a posterior probability of the adverse health event occurring within a predetermined period of time.

1 . A system for monitoring health events, the system comprising:
 an implantable medical device (IMD) comprising a plurality of electrodes and configured for subcutaneous implantation in a patient, wherein the IMD is configured to determine one or more subcutaneous tissue impedance measurements via the electrodes; and   processing circuitry coupled to the one or more storage devices, and configured to:
 determine a respective one or more values for each of a plurality of physiological parameters, the plurality of physiological parameters including one or more subcutaneous tissue impedance parameters determined from the one or more subcutaneous tissue impedance measurements; 
 identify a diagnostic state for each of the physiological parameters based on the respective values, the diagnostic states defining a plurality of evidence nodes for a probability model; and 
 determine, from the probability model, a probability score indicating a likelihood that the patient (a) is experiencing an adverse health event or (b) is likely to experience the adverse health event within a predetermined amount of time. 
   
     
     
         2 . The system of  claim 1 , wherein the determined values of the physiological parameters correspond to a preceding timeframe relative to when the probability score is determined. 
     
     
         3 . The system of  claim 1 , wherein the physiological parameters include values corresponding to at least one of: heart rate variability (HRV), night heart rate (NHR), patient activity (ACT), atrial fibrillation (AF), R-wave amplitude, heart sounds, or ventricular rate. 
     
     
         4 . The system of  claim 1 , wherein the processing circuitry is configured to:
 identify, from the respective one or more values for each physiological parameter, a plurality of physiological parameter features that encode amplitude, out-of-normal range values, and temporal changes; and   identify the evidence nodes based at least in part on the plurality of physiological parameter features.   
     
     
         5 . The system  claim 1 , wherein the probability model is a Bayesian Network comprising at least two child nodes and a parent node. 
     
     
         6 . The system of  claim 1 , wherein the processing circuitry is configured to:
 determine an input to a first child node of the plurality of evidence nodes based on the respective one or more values of the one or more subcutaneous tissue impedance parameters; and   determine an input to a second child node of the plurality of evidence nodes based on a combination of one or more values indicating an extent of atrial fibrillation (AF) in the patient during a time period and one or more values indicating a ventricular rate during the time period.   
     
     
         7 . The system of  claim 1 , wherein the probability model is expressed as:
 P(d, e 1 , . . . , e N )=P(d)Π i=1   N P(e i |d), wherein P(d) comprises a prior probability value, P(e i |d) comprises a conditional likelihood parameter, d comprises a parent node, and e 1 -e N  comprise the evidence nodes.   
     
     
         8 . The system of  claim 1 , wherein the processing circuitry is further configured to:
 compare the probability score to at least one risk threshold; and   determine one of a plurality of discrete risk categorizations based on the comparison.   
     
     
         9 . The system of  claim 1 , wherein the processing circuitry is further configured to:
 identify an occurrence of missing data, the missing data corresponding to a particular physiological parameter;   determine an extent to which the data for the physiological parameter is missing; and   determine whether to use the physiological parameter when determining the probability score based on the extent to which the data is missing.   
     
     
         10 . A method comprising:
 determining a respective one or more values for each of a plurality of physiological parameters, the plurality of physiological parameters including one or more subcutaneous tissue impedance parameters determined from one or more subcutaneous tissue impedance measurements;   identifying a diagnostic state for each of the physiological parameters based on the respective values, the diagnostic states defining a plurality of evidence nodes for a probability model; and   determining, from the probability model, a probability score indicating a likelihood that the patient (a) is experiencing an adverse health event or (b) is likely to experience the adverse health event within a predetermined amount of time.   
     
     
         11 . The method of  claim 10 , wherein the determined values of the physiological parameters correspond to a preceding timeframe relative to when the probability score is determined. 
     
     
         12 . The method of  claim 10 , wherein the physiological parameters include values corresponding to at least one of: heart rate variability (HRV), night heart rate (NHR), patient activity (ACT), atrial fibrillation (AF), heart sounds, or ventricular rate. 
     
     
         13 . The method of  claim 10 , further comprising:
 identifying, based on the one or more subcutaneous tissue impedance measurements, a periodic variation in subcutaneous tissue impedance; and   determining, based on the periodic variation in subcutaneous tissue impedance, a parameter value for at least one of the physiological parameters.   
     
     
         14 . The method of  claim 10 , further comprising:
 identifying a plurality of physiological parameter features based on the respective one or more values for each physiological parameter, wherein the features are configured to, upon analysis, yield a same number of potential diagnostic states for each physiological parameter; and   identifying, from the potential diagnostic states, the diagnostic state for each of the physiological parameters.   
     
     
         15 . The method of  claim 10 , wherein the probability model is a Bayesian Network comprising at least two child nodes and a parent node. 
     
     
         16 . The method of  claim 10 , further comprising:
 determining an input to a first child node of the plurality of evidence nodes based on the respective one or more values of the one or more subcutaneous tissue impedance parameters; and   determining an input to a second child node of the plurality of evidence nodes based on a combination of one or more values indicating an extent of atrial fibrillation (AF) in the patient during a time period and one or more values indicating a ventricular rate during the time period.   
     
     
         17 . The method of  claim 10 , wherein the probability model is expressed as:
 P(d, e 1 , . . . , e N )=P(d)Π i=1   N P(e i |d), wherein P(d) comprises a prior probability value, P(e i |d) comprises a conditional likelihood parameter, d comprises a parent node, and e 1 -e N  comprise the evidence nodes.   
     
     
         18 . The method of  claim 10 , further comprising:
 comparing the probability score to at least one risk threshold; and   determining one of a plurality of discrete risk categorizations based on the comparison.   
     
     
         19 . The method of  claim 10 , further comprising:
 determining, for each of the plurality of physiological parameters, the respective one or more values determined at various frequencies;   determining the diagnostic states using the respective one or more values; and   storing, to a memory device, at least one of: the respective one or more values or the probability score.   
     
     
         20 . A non-transitory computer-readable storage medium having stored thereon instructions that, when executed, cause one or more processors to at least:
 determine a respective one or more values for each of a plurality of physiological parameters, the plurality of physiological parameters including one or more subcutaneous tissue impedance parameters identified from one or more subcutaneous tissue impedance measurements;   identify a diagnostic state for each of the physiological parameters based on the respective values, the diagnostic states defining a plurality of evidence nodes for a probability model; and   determine, from the probability model, a probability score indicating a likelihood that the patient (a) is experiencing an adverse health event or (b) is likely to experience the adverse health event within a predetermined amount of time.